02 Corrosion Monitoring Manual

Corrosion Monitoring Manual A comprehensive guide to corrosion monitoring in oil and gas production and transportation facilities S Webster, R Woollam Sunbury Report No. ESR.95.053 dated November 1996

Main CD Contents

Contents Summary

1

Acknowledgements

3

Introduction to Corrosion Monitoring

5

Background Elements of a Corrosion Control Strategy What are Corrosion Monitoring Methods? The Economics of Corrosion Monitoring General Guidelines Selection of a Corrosion Monitoring Location and Technique Design of Corrosion Monitoring Location Process Monitoring Data Handling Side-stream Monitoring Corrosion Monitoring: A System by System Approach Critique of Corrosion Monitoring Methods Introduction Weight Loss Coupons Electrical Resistance Methods Linear Polarisation Resistance Method (LPR) AC Impedance Electrochemical Noise Galvanic Corrosion Monitors Hydrogen Probes Process Stream Monitoring Ultrasonic Thickness Measurement Radioactive Methods

5 6 8 10 15 15 22 31 34 36 37 45 45 45 50 60 64 66 70 72 76 82 84

References

87

Appendix 1

91

Appendix 2

93

Index

95

Summary

Lower cost materials are the natural economic choice for oil and gas production and transportation facilities. Unfortunately these materials (e.g. carbon steel, low alloy steels) in general have a low resistance to corrosion. Therefore, the corrosion risks of these materials have to be proactively managed. To this end BPX have developed and implemented corrosion control strategies which integrate risk assessment and corrosion control with corrosion monitoring and inspection. The application of corrosion monitoring as part of a corrosion control strategy is complex and often becomes the responsibility of engineers who are not experts in the field. This manual has been developed as an aide to those designing and operating a corrosion monitoring system. The manual focuses on techniques which are classically called corrosion monitoring techniques. The aim of the manual is not to be prescriptive or disregard conventional or other approaches/techniques but rather to put in place guidelines which will aid any operator concerned with corrosion monitoring. This document supplements the BP recommended practice on corrosion monitoring, RP6-1, by providing more in-depth information and advice based on recent operational experience. This manual addresses: ❍

Choice of monitoring location/orientation.

❍

Choice of monitoring technique.

❍

Application of the various techniques.

❍

Critique of monitoring techniques.

Other complementary methods such as inspection, intelligence pigging and downhole surveys are outside the scope of this manual and are covered by the BP recommended practice RP 32-4. Many of the principles and concepts given here are also relevant to processing facilities such as glycol and amine gas treatment systems, although these cases are not dealt with specifically. The main points in this Corrosion Monitoring Manual are summarised in a shorter companion document: S Webster, R C Woollam, “Corrosion Monitoring Guidelines”, Sunbury Report No. ESR.95.055, dated November 1996.

1

Acknowledgements

The authors would like to thank, ❍

BP staff who helped in useful discussions

❍

Suppliers who provided information and photographs

❍

Drew McMahon (ESS Sunbury) for editorial assistance

3

Introduction to Corrosion Monitoring

Background A 1988 survey revealed that BP transports 80% of its cash flow through facilities that are over 15 years old [1]. The integrity of such facilities is vital to the successful and profitable operation of the Company and the prevention of environmentally sensitive incidents. Although BP has a first class record in environmental issues, major pipeline repairs and replacements alone have cost BP around $250 million over the past 5 years. A recent survey [2] of BPX’s corrosion costs in the North Sea estimated that corrosion accounts for over 10% of the lifting costs per barrel of oil. Lower cost materials are the natural economic choice for oil and gas production and transportation facilities. Unfortunately these materials (e.g. carbon steel, low alloy steels) in general have a low resistance to corrosion. Therefore, the corrosion risks of these materials have to be managed proactively. To this end BPX have developed and implemented corrosion control strategies which integrate corrosion monitoring with risk assessment and corrosion control. The aim of corrosion monitoring is primarily to ensure that the design life is not being adversely affected or compromised and also to maximise the safe and economic operational life of a facility by: ❍

Safe operation of a process plant Corrosion can compromise plant integrity. If a plant is to be operated safely, all corrosion risks must be monitored.

❍

Improvement in the economic operation This aims to optimise corrosion control activities (e.g. corrosion inhibitor injection rates, oxygen concentrations, flowrates, etc.) whilst minimising operational costs.

❍

Improvement of maintenance/shut-down scheduling This is having sufficient knowledge of the plant condition and accurate life prediction to avoid unscheduled shutdowns due to unforeseen failures. Also to ease inspection load during planned shutdowns and to optimise spares stocks.

5

INTRODUCTION TO CORROSION MONITORING

❍

Assessment of impacts of process/operational changes and upsets This aims to obtain information on the relationship between corrosivity, process and operating variables such as flow rate, water cut, temperature, etc. In addition to these are the effects of corrosion control methods (corrosion inhibition, increased corrosion allowance, corrosion resistant alloys) and process changes and upsets (one-off well workovers such as acidisations well intervention, onset of water breakthrough, introduction of third party fluids).

Elements of a Corrosion Control Strategy There are three main components in the development and implementation of the BPX corrosion control strategies. Figure 1: A Schematic of the Inter-Relationships in a BPX Corrosion Control Strategy.

Risk / Criticality Assessment

Control Procedures

❍

Inspection and Monitoring

Risk/criticality assessment Risk assessment involves the identification of the main corrosion mechanism(s) possible for a given material/fluid combination, and the consequences of such corrosion occurring. Criticality assessment combines the consequences of such failure with the probability of it happening. The assessments can be undertaken in many ways but there should be a well-defined

6

INTRODUCTION TO CORROSION MONITORING

auditable trail (e.g. via a proprietary criticality assessment [3, 4]). Individual corrosion mechanisms depend on specific parameters that can be controlled and monitored in different ways. In most cases the probability of internal corrosion is dominated by only one or perhaps two corrosion mechanisms. ❍

Control procedures Once the main risks have been identified and assessed, measures to mitigate the risks need to be identified, detailed and implemented. In many cases the predicted/measured rates of attack may be acceptable and it is sufficient to keep the key parameters within their design limits. In other cases additional measures will be required, e.g. material selection or chemical treatment (corrosion inhibitor, oxygen scavenger, etc.).

❍

Monitoring and inspection In all cases monitoring and inspection procedures will have to be put in place to confirm: ❍ ❍ ❍

Actual vs. predicted corrosion rates Process parameters within design limits Correct operation of control measures

Monitoring and inspection are two overlapping tasks. The first is the ongoing monitoring of the corrosion process and the measures taken to control it. The second is the provision of mechanical integrity assurance. Inspection also provides datum points against which corrosion monitoring is often related or quantified. In a corrosion control strategy these tasks aim to determine whether the expected corrosion is actually occurring, the corrosion rate, and the effectiveness of any control measures. Figure 2 classifies currently available inspection and monitoring techniques indicating the complementary characteristics of each technique.

7

INTRODUCTION TO CORROSION MONITORING

Figure 2: Classification of Currently Available Inspection and Monitoring Techniques

➚ ➚ ➚ ➚ ➚ ➚

Ultrasonics

Flexible UT Mats

Auto UT

Coupons

Field Signature Method (FSM)

Electrical Resistance Probes

Chemical Analysis

Linear Polarisation Resistance

Electrochemical Noise

Small time interval between reading Destructive (probes / consumables) Indirect measure of material loss High sensitivity Lower accuracy / reliability Leading indicator

Radiography

Large time interval between readings Non-destructive Direct measure of material loss Low sensitivity High accuracy / reliability Lagging indicator

Visual Inspection

➚ ➚ ➚ ➚ ➚ ➚

✓ ✓

✓ ✓

✓ ✓

✓ ✗

✓ ✓

✓ ✓

✓ ✓

✓ ✗

✓ ✗

✓ ✗

✓ ✗

Uniform Corrosion Localised Corrosion

Notes: (a) the position of a technique in the table does not relate to its exact position along the arrows (b) Flexible UT mats maximum temperature is 120°C

All of the above activities (risk assessment, corrosion control inspection and monitoring) are interdependent. Results from corrosion monitoring and inspection must be used to re-evaluate and modify, where necessary, the risk and criticality assessment and any control procedures. This report focuses on the corrosion monitoring elements in a corrosion control strategy.

What are Corrosion Monitoring Methods? When undertaking corrosion monitoring it is important not to rely on just one method. The best results are obtained by using a range of techniques. Corrosion monitoring in this context can be defined as:

The use of any method that enables an operator to estimate or measure the corrosion rate occurring in service of an item of plant, or the corrosivity of a process stream.

\ The main methods fall into the following four categories:

8

INTRODUCTION TO CORROSION MONITORING

Inspection These techniques are used to assess wall thickness changes, and detect material defects with the possibility of detecting pit growth or crack propagation. The techniques most commonly used are: ultrasonics, magnetic flux (on-line inspection vehicle), radiography, acoustic emission, thermography, visual examination, dye penetrant and magnetic particle inspection. On-line corrosion monitoring These techniques are used to assess changes in corrosivity with time. Typically the techniques are probe based and include electrical resistance methods, electrochemical methods and weight loss coupons. Analysis of process streams This approach is the monitoring of key process variables that affect stream corrosivity. i.e. pressure, temperature, production rates, fluid composition, production chemistry laboratory data (bio-activity, pH, oxygen content, chlorine, etc.) , corrosion product concentration (Fe or Mn concentration) and chemical treatments (dose rate and frequency). Many of these methods have fast response times ([O2], pH, etc.) and are used to monitor process control. A good example is the use of on-line oxygen monitoring to maintain an acceptable oxygen content in a sea water injection stream to control corrosion rates. Process stream data can be used with mathematical models to predict the potential corrosion rates throughout the facility. However, the main value of process stream data is for ensuring that any control activities are working and that when corrosion has been detected effective data analysis can be undertaken to identify the cause. Operational history assessment This approach is the analysis of previous data as an aide to providing information about the present and predicted corrosion rates. This includes:

9

INTRODUCTION TO CORROSION MONITORING

❍

Examination of production and operational records, including details of process changes (or upsets) which can give an insight to the corrosivity of the system.

❍

Failure analysis and inspection data can be used to predict parts of a facility most susceptible to certain modes of attack and subsequent failure.

Historically, corrosion monitoring and process data analyses were perceived as quite separate from inspection activities. Although inspection has been historically concerned with mechanical integrity, many inspection techniques can be used as corrosion monitoring tools. The complementary nature of these approaches is summarised in Figure 2. For any corrosion monitoring/inspection programme to be fully effective it is vital that all of the above information can be accessed centrally and compared together. This can be achieved by ensuring full access to all databases which hold the relevant information and having the appropriate software to conduct the relevant correlational analysis. The guidelines in this report focus on techniques which are classically called corrosion monitoring methods. Corrosion monitoring aspects are summarised in the current BP recommended practice RP6-1. This report supplements RP6-1 and provides a practical guide to corrosion monitoring giving full details on the design and application of a corrosion monitoring system. The aim is not to be prescriptive or disregard conventional inspection techniques but rather to put in place guidelines which will aid any operator concerned with corrosion monitoring.

The Economics of Corrosion Monitoring In general the purpose of corrosion monitoring is to optimize corrosion mitigation/repair/replacement activities such that an optimum between corrosion control and replacement costs is achieved. It should be noted that there may be additional cost considerations related to safety, environmental and production impacts which are NOT considered in the following. A given corrosion monitoring method or technique has only a limited accuracy and therefore, each corrosion rate determination has a

10

INTRODUCTION TO CORROSION MONITORING

random error associated with it. This random error can only be reduced by increasing the amount of corrosion monitoring undertaken but this will increase the overall costs of the activity. For an optimal corrosion monitoring program the benefit obtained should be greater than the cost incurred. For corrosion inhibitor optimisation there is a trade-off between replacement costs and corrosion inhibition costs (Figure 3a). This results in an operational minimum of the sum of the corrosion inhibition costs and the pipeline equipment repair or replacement costs. In order to determine the optimum corrosion inhibitor injection rate, the corrosion rate for the system needs to be determined. The corrosion rate will determine if the corrosion inhibitor injection rate is effective, if it needs adjustment (up or down), or if some alternative means of control is required (e.g. corrosion resistant alloys, CRAs) Figure 3a Shows the Trade-off Between Replacement Costs and Inhibition Costs.

Total Cost: CI and Replacement

Total Cost Replacement cost Inhibition cost

0

5

10

15

20

25

30

35

40

45

50

Corrosion Rate, mpy

Figure 3b Shows the Increasing Confidence and Reduction in Error, in Determining the Corrosion Rate as the Number of Corrosion Rate Measurements is Increased.

Corrosion Rate Spread with # Locations

2 5 10 20 50 100

Total Pv ∆ CR with #

0

5

10

15

20

25

30

35

40

45

50

Corrosion Rate, mpy

11

INTRODUCTION TO CORROSION MONITORING

Figure 3c The Increasing Cost of Corrosion Monitoring as the Number of Measurements Increases.

Cost/Benefit of Monitoring

Cost/Benefit of Monitoring Cost of Monitoring Program per Year

0

0.5

1

1.5

2

2.5

3

3.5

4

Log (Number of Locations)

Figure 3c clearly shows the point at which no additional corrosion monitoring is warranted as the incremental savings from corrosion inhibitor optimisation are less than the cost of the monitoring program. The cross-over of the two curves indicates the level of corrosion monitoring required to optimize the overall cost structure. This approach can be generalised to other corrosion mitigation methodologies and the monitoring of these systems. In general there is an optimum amount of corrosion monitoring in a system above which the costs of monitoring exceed any savings generated. Costs

12

Table 1 gives outline costs for various corrosion monitoring techniques. This table is a guide to the relative costs of each technique (hardware) and any operational costs associated with installation and data analysis. The costs will vary depending upon asset location and number of monitoring locations. However, these figures are a guide to the costing of monitoring/inspection activities.

INTRODUCTION TO CORROSION MONITORING

Table 1: Outline Costs for Various Corrosion Monitoring Techniques based on 1995 Information (£1 = $1.6).

Monitoring Method

Hardware

Probe

Man-hour costs

Weight Loss Coupons

None

£300

Coupon insertion and retrieval. Coupon analysis

Electrical Resistance probes

£1500

£500

Probe insertion and retrieval. Data analysis

Electrical Resistance Sand Monitor

£25000

£1000

Probe insertion and retrieval. Data analysis

FSM (Topsides)

£30000

-

Data analysis

FSM (subsea)

£250000

-

Data analysis

LPR

£1500

£300

Probe insertion and retrieval. Data analysis

Electrochemical Noise

>£2500

£300

Probe insertion and retrieval. Detailed data analysis. Very time consuming

Flexible UT Mats

>£2500

£300-600

Data analysis

13

General Guidelines

Selection of a Corrosion Monitoring Location and Technique

Introduction

The selection of the appropriate monitoring location(s) and technique(s) is critical for successful corrosion monitoring. It cannot be stressed enough that selection of the wrong location or technique will result in a large amount of effort and expense only to generate inappropriate or even misleading information. In many cases incorrect selection is worse than no selection as the quality of data are often not questioned. Physical access is important but should not dictate monitoring location. However, when a monitoring point is identified the position should allow routine access for probe maintenance, retrieval etc.

Incorrect selection of location or technique is worse than no selection.

All corrosion monitoring (and inspection) locations and methods must be recorded on the relevant technical drawings. This should include process flow diagrams, process and instrumentation diagrams (P&ID’s) and the isometric diagrams (PFD’s). On new facilities they should be included in the Computer Aided Design (CAD) system as this aids data analysis and the development of control procedures. The records should include not only details on the system, item and location, but also the method and probe orientation. There are no fixed rules on how to select a corrosion monitoring location or technique but the first step must be to decide the types of corrosion mechanisms to be monitored. Experience has shown that the following approaches are of value. Approaches to Selection

❍

Historical approach Experience at other assets utilising similar facilities is often the best source of advice regarding the most suitable locations and/or monitoring techniques. Inspection/shutdown reports and maintenance lists can provide valuable information on which parts of a facility have experienced the most severe corrosion.

15

GENERAL GUIDELINES

This aspect is vitally important at the design stage where operator feed back could prevent costly mistakes being re-made and minimise the cost of subsequent retro-fitting. Design contractors have limited operational experience and so it is important that BP assets support this activity by providing feedback and lessons learnt. ❍

Inspection/corrosion monitoring data Operating assets can provide valuable information by utilisation of existing inspection/corrosion monitoring data to identify the most suitable locations for future monitoring/inspection.

❍

Networking A wide range of disciplines need to be networked to obtain a full picture of current and potential future problems. For example production engineers can provide information on production profiles and well intervention programmes which may influence corrosion; production chemists have knowledge on fluid properties and chemical control measures which may influence corrosion; maintenance engineers can identify where most failures or replacements have been located. Two examples from recent BP operations are given here and highlight where inappropriate selection of the monitoring location or technique caused problems:

Corrosion monitoring in a main oil export line with water cut below 1%. Corrosion monitoring was undertaken using an intrusive electrical resistance probe via a top of the line access fitting. Low corrosion rates were observed which appeared to be insensitive to process changes. In this case the probe response was most likely reflecting the corrosivity of the continuous hydrocarbon phase and not that of the aqueous phase which constituted the corrosion hazard. A more reliable approach may well have been to have used a flush mounted electrical resistance probe via an access fitting located at the bottom of line where water separates out.

16

GENERAL GUIDELINES

Corrosion monitoring in a sulphide containing produced water line. Corrosion monitoring was undertaken using a flush mounted linear polarisation resistance (LPR) probe via a bottom of line access fitting. The monitoring programme yielded an exponentially increasing corrosion rate with time. In this case the most likely explanation of the results was that the probe response reflected the shorting out of the probe elements due to the formation of a conducting sulphide film. A more reliable approach may have been to substitute the LPR probe with a flush mounted electrical resistance probe in the same location. Selection of Location within Plant for Corrosion Monitoring

Figure 4: A Check List for Identifying a Corrosion Monitoring Location

This section outlines the main points that should be considered when identifying a corrosion monitoring location. These are summarised in Figure 4. TOPIC

CONSIDERATIONS Single or multiphase flow

What is the major corrosion mechanism and mode of attack ?

Corrosion rate of each phase Mechanism/mode of attack

Prior elevation changes

Orientation of pipework?

Water drop out Other pipeline entrants Position of other pieces of equipment

Location of chemical injection points

Upstream/downstream effects Localised effects Corrosivity of injected chemical

Environment indicative of corrosion elsewhere

Low alloy probe in CRA line

Identify process changes in system

Process changes Location reflect most corrosive situation

Physical access

Should not dictate locations

Record locations

Flow diagram process and instrument diagrams etc.

17

GENERAL GUIDELINES

Direct Monitoring

From the initial criticality assessment the predicted internal corrosion rates will have been identified. However, the following factors need to be considered when selecting the most appropriate monitoring points in a given system. ❍

Location of corrosive phase From the predicted corrosion rates the most likely location for corrosion to occur for a given phase must be identified. A good example is the transportation of wet gas. In this case corrosion related to water drop out will occur at the bottom of line. Corrosion at the top of line will occur as a result of water condensing from the gas phase.

❍

Mode of attack The anticipated corrosion mechanisms and modes of attack must be understood (general or localised attack, stress corrosion cracking, under deposit corrosion, process upset detection etc.) This will determine the siting of any monitoring points and help in selecting the most appropriate monitoring techniques at those points.

❍

Flow effects The flow rate and flow regime has a major impact on corrosivity and the location of the attack. The current BP corrosion monitoring recommended practise RP 6-1 [5] states that "access fittings should be located a minimum distance of 7 pipe diameters downstream of and a minimum of 3 pipe diameters upstream of any changes in flow caused by bends, reducers, valves etc". This is to ensure that the probe is sited in a region where water drop out is more likely. This location also ensures that the hydrodynamics are more uniform and so will provide a fluid corrosivity represenative of most of the pipe. Higher or lower corrosivities are possible in the hydrodynamically severe regions such as bends, reducers, valves, elevation changes and areas close to some major pieces of equipment (eg pumps). For example, BPX Alaska have had accelerated corrosion at road crossings (multiple elevation changes), and Magnus have had accelerated corrosion in the tortuous discharge pipework from the main-oil-line booster pumps. This acceleration can arise through enhanced water drop-out and wetting via centrifugal action (Figure 5). In

18

GENERAL GUIDELINES

Figure 5: Effect of Elevation Change on Water “Drop Out”

CORROSION MONITORING LOCATIONS DOWN STREAM OF EXPANSION Not to scale

Water Accumulation

Corrosion monitoring locations

0

2

4

6

8

10

12

14

Pipe Diameters

CORROSION MONITORING LOCATIONS AFTER ELEVATION CHANGE

CORROSION MONITORING LOCATIONS AFTER ELEVATION CHANGE Wet Gas Low Velocity

Oil / Water Low Velocity 80:20 Oil / Water

Flow

Gas

Water

Flow

Corrosion Monitoring Location

Oil

Water

Corrosion Monitoring Location

Flow

Gas

Water

Flow

Oil

Water

19

GENERAL GUIDELINES

contrast, in water/crude oil systems it is equally possible to have reduced corrosivity at elevation changes because the extra turbulance causes emulsification of the water. Water hold-up and water drop-out effects are therefore of central importance in deciding the optimum corrosion monitoring location. Water hold-up, drop-out and deposit build up are all less likely in vertical sections than in horizontal. Drop-out is most likely in long horizontal pipe runs. ❍

Process stream changes Process changes (pressure, temperature, flow rate etc.) will affect potential corrosivity due to solution chemistry changes. It is important to consider process changes in the system to ensure that the chosen monitoring location coincides with the location of highest corrosivity. The position of equipment which affects process conditions (e.g. vacuum/gas stripping towers in sea water systems, pumps, heat exchangers) should also be considered. Third party entrants: Consideration should be given to other entrants to a pipeline system as these could influence corrosivity considerably. This may include mixing of separate well streams, through to third party entrants from other fields. Factors which are important include: water cut, flow rate, inhibition levels, water chemistry effects (pH, scaling), CO2 and volatile fatty acid content. Location of chemical injection points: The injection of production chemicals (corrosion inhibitors, scale inhibitors, demulsifiers, oxygen scavengers, etc.) can have a marked effect on corrosion. It is important to consider the positions of injection points when siting corrosion monitoring locations. In some cases it may be pertinent to monitor both upstream and downstream of the production chemical injection point. Scale inhibitors can be corrosive to certain steels and render corrosion inhibitors less effective if they are not fully compatible.

BPX Alaska have experienced high corrosion rates close to scale inhibitor injection points due to poor positioning of the injection quills. This aspect of localised attack is of importance to integrity inspection.

20

GENERAL GUIDELINES

Indirect Monitoring

❍

Environment indicative of corrosion elsewhere In areas of high corrosivity, corrosion resistant alloys (CRAs) are often specified. Locating a carbon steel coupon/probe in a CRA line can give important data regarding the potential corrosivity of the processed fluids to other carbon steel equipment located downstream or in other parts of the system. A good example is the use of CRAs in the produced water lines from a first stage separator. Small amounts of residual produced water will enter the main crude oil export line after further processing. Monitoring the corrosivity of the fluids in the CRA line will give information on the potential corrosion rates in the carbon steel export line. However, extrapolation is needed to ensure system changes are taken into account.

❍

Process stream monitoring This approach is the monitoring of key process variables that affect corrosion rates. The measured variables - pressure, temperature, production rates, fluid composition, production chemistry data (bio-activity, pH, oxygen content, chlorine etc.), corrosion product concentration (Fe or Mn concentration) and chemical treatments (dose rate and frequency) - can be used with predictive models and current corrosion knowledge to give a reasonable estimate of potential corrosion rates.

Corrosion Monitoring Technique Selection

This section covers the selection of a corrosion monitoring technique. Details of inspection-based activities are given in the BP Standard RP 32-4 which covers: ❍ ❍ ❍

Inspection scope and frequency Inspection techniques Inspection pigging technology

The inspection or monitoring technique selected must provide information relating to the actual corrosion mechanisms. Consideration of the corrosion environment is important as this will often preclude many techniques (e.g. electrochemical methods are not suitable in low water cut or low conductivity situations). An understanding of the anticipated corrosion mechanisms (general, pitting, cracking etc.) is also important as this will give an insight into

21

GENERAL GUIDELINES

the most suitable monitoring technique(s) and eliminate many that are unsuitable (e.g. pitting most easily detected using weight loss coupons). The application of any of the techniques must be carefully considered (see Data Handling p34) because the economic benefit must outweigh the cost of the activity. A schematic for selecting a corrosion monitoring method(s) is presented in Figure 6. There are no fixed rules on which methods are most suited for a given system (i.e. water injection system, crude oil flow lines etc.) as the conditions in each can vary. However, Table 2 gives a general guide to the possible application of the various monitoring techniques on a system by system approach. When selecting a monitoring technique it must be realised that each technique gives only a limited amount of information. It is good practice to use a selection of techniques to give overall confidence in the results. The first choice must always be inspection-based methods as they are very reliable for integrity assurance. This can then be supported by probe-based methods. If only one probe-based method can be used then the first choice should be weight loss coupons as this technique gives both general and localised information. The utility of inspection based methods is tempered by the fact that they are “lagging” indicators of corrosion. If inspection data says the situation is bad then it may be too late to do anything about it because the damage has already been done. Monitoring methods are “leading” indicators of corrosion. They show the fluid corrosivity at a particular moment, potentially before any significant damage has occurred. Hence, monitoring methods are always a valuable complement to inspection methods.

Design of Corrosion Monitoring Location Access Fittings

22

The insertion of probes and coupons into pipework and facilities without the need for plant shutdown relies on the use of proprietary

GENERAL GUIDELINES

Figure 6: Schematic for Selecting a Corrosion Monitoring Method

23

Electrical Resistance Probes Electrochemical Probes

Galvanic Probes

Bacterial Monitoring

May be useful for condensed water in wet gas export line

7

✓ ✓ ✓ ✓ ✓ ✓ ✓

✓ ✓ ✓ ✓ ✓ ✓ ✓

Maximum temperature 120°C

✗ ✓ ✗ ✓ ✓ ✗ ✗

6

✓ ✓ ✓ ✗ ✗7 ✓ ✓

Intrusive probe prefered. Flush mounted unsuitable where biofilming tendency.

H2S

O2,H2S

✗

CO2,H2S

CO2,H2S

CO2,H2S

O2,CI2

5

✓ ✓ ✓ ✗ ✗ ✓ ✗

Depends on water quality. LPR unsuitable where biofilming tendency.

✓ ✓ ✓ ✗ ✗ ✗ ✗

4

Weight Loss Coupons / Spool Pieces

Only in water cuts above ca. 10%-20%.

✓ ✗ ✓ ✗ ✗ ✓ ✓

Suspended Solids

3

✗

2

✓ ✗ ✗ ✗ ✗

Dissolved Solids

May be used where oxygen content is high.

✗ ✗ ✗ ✗

1

✓ 3

✓4

Dissolved Gases

2

✓5 ✓ ✓ ✓ ✓ ✓ ✓

pH

Depends on water quality. LPR unsuitable where there is a low ion content or a strong scaling tendency (or other form of electrode contamination is possible).

✓ ✓ ✓ ✓ ✓ ✓ ✓

Hydrogen Probes / Patch

1

Storage Vessels with Separated Water Bottom

Effluent Water

Hydrocarbon Gas

Unstabilised Crude Oil

Aquifer Water

Corrosion Product Analysis

✓ ✓ ✓ ✓ ✓ ✓ ✓

6 Flexible UT mats

24 Field Signature Method

Table 2: A General Guide to the Application of Corrosion Monitoring Techniques

Flow Lines (oil, water, gas)

Seawater Injection

GENERAL GUIDELINES

GENERAL GUIDELINES

two inch access fittings. Access fittings are usually installed at the construction phase or during subsequent planned plant shut downs. Therefore, it is important that the corrosion monitoring requirements are well thought out at the design stage otherwise subsequent installation can be difficult and costly. Access fittings can be installed during plant operations using a hot tap but in many circumstances safety concerns will preclude such activity. If this is the case, there could be a long time interval before information is gained. Access fittings are suitable for operating pressures up to 137 bar(g) (2000 psi(g)). Typical high pressure access fittings are shown in Figure 7. Figure 7: Typical High Pressure Access Fitting Design

Flareweld

Flareweld Flanged Tee

Flareweld Threaded Tee

Flange

The BP recommended practice RP 6-1 gives a thorough overview of access fittings and retrieval tools. However, the following points should also be considered.

25

GENERAL GUIDELINES

❍

Orientation of access fitting RP6-1 recommends top of line access fittings. This general statement can be misleading and often the location should be at other orientations. The bottom of line location can cause problems with accumulation of debris and the possible galling of the threads. However, modern access fittings (e.g. CorrOcean hydraulic access fitting) or improved retrieval procedures can minimise these effects. BPX Norway have developed a procedure to minimise the effect of debris build up on bottom of line location using standard access fittings [6]. This involves back pressuring the retrieval tool so that any debris is pushed back into the line. This procedure has been used on Ula since 1986 without any problems.

If corrosion is occurring at the bottom of line (e.g. wet oil, wet gas) then the access fitting should be located in this position. This will also minimise probe length and so reduce the possibility of fatigue failure and make the line easier to pig. The access orientation is less critical for a single phase water stream since all parts of the pipewall well experience the same environment. Retaining adequate clearance for the retrieval tool is important when locating an access fitting for a corrosion probe : otherwise the fitting will be unusable. RP 6-1 gives full details on the clearance required for different retrieval tools. However, it is vital that a fitting is not simply located at a particular point because it happens to have a convenient space. The location must be also capable of providing useful information or else it is not worth having. Of course, retrofitting a probe at the optimum corrosion monitoring location may be impossible in a mature plant. This emphasises the importance of good early corrosion engineering design. ❍

Access fitting and sampling point design For systems operating at pressures below 10 bar(g) (150 psig), low pressure access fittings can be used. Full details are given in RP 6-1. All probe and coupon holders used in low pressure

26

GENERAL GUIDELINES

fittings should be fitted with a blow out preventer to limit the slide out of the monitoring device during installation and retrieval. Safety clamps should also be used to secure retractable probes and coupon holders whilst on line. For operating pressures from 10-137 bar(g) (i.e. 150-2000 psig) proprietary 2” high pressure access fittings should be used. All access fittings should be fitted with heavy duty covers to protect the fitting threads and electrical connections from damage. The cover should be fitted with bleed plugs (or possibly a pressure gauge) so that any leaks which may have occurred between the access fitting body and the monitoring device can be easily identified. The design should not allow the probe to be inserted back-to-front. Sample points for the collection of process fluids should include two isolating valves in series, one of which should be a needle valve. Details are given in RP 42-1 [7]. ❍

Material selection Access fitting material should conform to the requirements of the piping specification. The welding of access fittings onto any equipment must comply with the requirements of the various codes. The solid and hollow plugs used in the access fitting should be manufactured from corrosion resistant material, with the choice of material depending on the service duty. Austenitic stainless steels are suitable for most carbon steel access fittings. The risk of thread galling rules out stainless steel plugs for corrosion resistant access fittings made from materials such as 316 and duplex stainless steel. Alternative corrosion resistant plug materials have been tried including a hard stainless steel called Nitronix 60 and a ceramic coated stainless steel. Both still suffer galling problems and are not recommended. The preferred material is to use carbon steel plugs which have been phosphate coated. This approach dictates that there is regular servicing of the access fitting to monitor the condition of the carbon steel plug. The problem of thread galling can be eliminated by converting existing threaded access fitting to hydraulic retrieval by means of a permanent adaptor. This approach has been taken by BP Norway on both the Ula and Gyda offshore production facilities. Furthermore, new projects should consider using hydraulic access fittings where access fittings are used on corrosion resistant piping materials.

27

GENERAL GUIDELINES

Recent experience at a refinery has highlighted the importance of material selection. A 316 stainless steel probe holder on a Crude Distillation Unit top pump-around-circuit suffered chloride stress corrosion cracking in service leading to a hydrocarbon leak and a serious “near miss”. The incident report recommended that all future fittings should be constructed from Hastelloy for this application in which chloride ion concentration and low pH put conventional austenitic stainless steels at risk.

It is also important that any seals associated with the probe assembly should have satisfactory performance under the operating conditions[8]. ❍

Trap-type monitoring point In low water cut situations some operators have adopted the use of water traps. These traps act as a sink for water drop-out and allow conventional monitoring methods to be employed. This approach is shown schematically in Figure 8. SECTION THROUGH PIPELINE SHOWING SIDE ENTRY

Figure 8: Schematic of Water Trap for Corrosion Monitoring

Pipe Wall

Product

Water Drop-out

Trap Assembly

Area required for retrieval

Probe

Drain Valve

28

GENERAL GUIDELINES

In this monitoring system the results reflect the inherent corrosivity of the fluid but do not allow other process conditions such as flow effects to be simulated. This is a major limitation. The trap can also promote bacterial activity which might not be typical of normal operations under flowing conditions. The design of the trap must include appropriate isolation to allow accumulated water to be drawn off. Traps can also be used to collect water samples from low water cut situations for laboratory evaluation. Traps can become a potential corrosion site by acting as a dead leg and so their use is not recommended. Probe Configuration

Probes and coupons fall into two main categories: ❍

Flush mounted These are designed to be positioned so that the probe element is flush with the inside pipe wall. This approach will simulate processes which occur at the pipe wall surface. Typically this type of probe would be used to monitor corrosion in low water cut situations (e.g. wet oil), water drop out (e.g. wet gas), under deposit corrosion, and areas where water condenses. A turbulent location will help reduce fouling.

❍

Intrusive These probes protrude well into the process stream and are suited for measuring the overall corrosivity of a process stream rather than specific aspects like the flush mounted probe. Typically they are used to monitor process upsets in a single phase, high wall shear stresses, or “worst case” situations. They are especially well suited to clean water streams (e.g. sea water injection). In dirty systems (bugs, suspended solids etc) they are less likely than intrusive probes to become fouled, especially if they are in a turbulent location. However, they cannot be used in lines which are pigged.

There is no generalisation as to which of these probe configurations is most appropriate for corrosion monitoring. The choice will depend on the information required. Validation of Monitoring Method Response

Before embarking upon a detailed corrosion monitoring programme it is important to ensure that the response of the monitoring method is sufficiently sensitive and reliable and responds to changes in the conditions being monitored. 29

GENERAL GUIDELINES

This aspect is often overlooked when undertaking routine monitoring. If data validation has not been undertaken, corrosion monitoring data can actually be misleading. This can lead to complacency in corrosion control, unnecessary modification of control methods or changes to operational parameters being made. Increasing probe corrosion rate is usually a warning of increasing corrosivity but a low probe corrosion rate is not a guarantee that a system is under control.

A lack of probe response has often been interpreted as a sign of good control rather than a sign of poor positioning/choice of the monitoring technique.

For any monitoring programme control checks must be included to ensure the reliability of the data. These should include routine cross checks with other methods, checking of process data for major changes in operating conditions that should lead to changes in corrosivity (e.g. increase in water cut) or deliberately changing the corrosivity of the system and monitoring the probe response. When changing the corrosivity of the system the full risks of the operation must be identified, including the fact that the data being collected could have limited value. A recent corrosion survey [9] of a sea water injection system showed good oxygen and free chlorine control according to the on-line dissolved gas monitors. However, when the responses of the on-line monitors were checked against proprietary chemical kits the levels of dissolved gases were found to be an order of magnitude higher than monitored. Also the probe responses were very slow, i.e. hours, to respond to instantaneous changes in dissolved gas levels. It was found that the probes were fouled and needed more frequent maintenance.

In the case cited above the electrochlorinator output was being adjusted based on poor quality on-line monitoring information. This resulted in an increase in hypochlorite concentration and a corresponding increase in corrosivity. In this case, process changes are now only made after confirmation of the on-line data by the manual chemical kits

30

GENERAL GUIDELINES

Process Monitoring Background

Process monitoring is a key aspect of any corrosion monitoring programme and covers a wide range of activities including but not limited to, the following: ❍

Measurements of standard process data (temperature, pressure, flow rates, water cuts etc.)

❍

Chemical analysis of the process streams (dissolved ions, bacterial levels, suspended solids, dissolved gases etc.)

❍

Chemical analysis of corrosion products.

❍

Details of production engineering activities (workovers, acidisations etc.)

❍

Monitoring the addition of production chemicals.

Full details of these activities are given in the Process Stream Monitoring section (p76). All of the above activities can have a major impact on corrosivity. Process monitoring is essential in predicting potential corrosivity and in the interpretation of corrosion monitoring data to validate the ongoing inspection/monitoring programme. Process monitoring measurements can be made either on-line or by samples taken at regular intervals. Sampling must be carried out correctly and the time and place recorded so that the data can be compared to other process monitoring and on-line corrosion monitoring information. In many cases the analysis of the sample can be undertaken at the site rather than in the controlled environment of the laboratory. This route is often preferred as it minimises the effect of sample ageing. There are standard procedures available for most of the methods discussed, the details of which are outside the scope of this document and are listed below.

31

GENERAL GUIDELINES

Figure 9: The Use of Process Monitoring Data to Predict Corrosion Risk

CO2

H2S

O2

O2

MIC

1 2 3 4 5 6 7 8 9 10 11

32

System

Oil & Gas Systems

Oil & Gas Systems

Sea Water

Injection Water

Oil & Gas Systems

Process Parameter Temperature Pressure Flow Regime Flow Rate CO2 mol % H2S mol % Water Chemistry pH Solids Corrosion Inhibitor - dose rate - deployment Temperature Pressure CO2 mol % H2S mol % Water Chemistry pH

Temperature Pressure Flow Rate O2 Free CI2 Biocide Oxygen Scavenger pH Water Chemistry Temperature Pressure Sessile Bacteria Count Planktonic Bacteria Counts pH Water Chemistry

Model

de Waard & Milliams (1) Cormed (2) pH Calc (3) Design Guidelines - Wet Gas (4) - Wet Oil (5) - Erosion (6)

NACE MR-0175 (7) Cormed (2) Design Guidelines - Material Selection (8)

Oldfield &Todd (9) Design Guidelines - Sea Water Injection (10) - Material Selection (8)

Microbiologically Influenced Corrosion Review (11)

C de Waard et. al., Prediction of CO2 Corrosion of Carbon Steel, NACE 93, Paper 69 1993 J L Crolet, "Cormed Lotus 123 Spread Sheet for Calculating pH of Produced Waters" Elf Aquitane-SNEA (P) Copyright 1988, 1990 pH prediction J Pattinson et. al., A Corrosion Philosophy for the Transport of Wet Hydrocarbon Gas Containing CO2, ESR. 93. ER016 J Pattinson et.al., A Corrosion Philosophy for the Transport of Wet Oil and Multiphase Fluids Containing CO2, ESR. 93.ER013 J Pattinson, Erosion Guidelines, ESR.94. ER070 NACE MR-0175 J Martin, Guidelines for Selecting Downhole Tubular Materials with Particular Reference to Sour Conditions, ESR. 94.ER043 Oldfield et.al., Corrosion of Metals in Dearated Seawater, BSE-NACE Corrosion Conference, Bahrain, Jan 19-21 81 J T A Smith, Minimising Corrosion of Carbon Steel in Sea Water Injection Systems - Guidelines for Water Quality, ESR. 94.005 I Vance, Microbiologically Influenced Corrosion (MIC) in Oil Production Operations, Topical report No 8615 1993

GENERAL GUIDELINES

API RP 38

Recommended Practice for Biological Analysis of Subsurface Injection Waters.

API RP 45

Recommended Practice for Analysis of Oilfield Waters

NACE RP 0173 Recommended Practice: Collection and Identification of Corrosion Products NACE RP 0192 Recommended Practice : Monitoring Corrosion in Oil and Gas Production with Iron Counts

The process data required to supplement the corrosion monitoring will depend on the application. However, consideration of the mechanism or mode of attack will help identify the most appropriate methods.

Application of Process Monitoring Data

The application of process monitoring data with inspection and corrosion data is important if the full value of the data is to be obtained and to provide the necessary assurance of plant integrity so minimising shutdowns and extending vessel inspection intervals. Typically the data will be used to assess corrosion rates indirectly (e.g. iron counts) or be used to predict potential corrosivity from a detailed knowledge of the corrosion processes. The latter approach is extremely valuable as it enables an operator to predict changes in corrosion rates and modify any monitoring/inspection activities or control procedures before significant damage has occurred. Figure 9 gives a broad summary of the use of process monitoring data.

33

GENERAL GUIDELINES

The Figure below is from the BP Magnus platform in the North Sea. A number of production vessels on the installation are sandwashed every day. It was assumed that all the water was routed out from the vessels via the drains. The fluid corrosivity graphs for the main-oil-line (MOL) generally showed a low value, however, high corrosion "spikes" were appearing every day. Comparing the times of these "spikes" with process conditions revealed that some of the sandwash water was in fact going down the MOL and increasing the corrosion. By doubling the corrosion inhibitor injection rate into the MOL during sandwashing, the "spikes" on the probe output disappeared and the normal low corrosion rate could be maintained.

Corrosivity (mm/yr)

0.10

Corrosion Peaks Related to Sand Washing 0.08 0.06

0.04

0.02

0.00

Days

Data Handling

The methods used to analyse corrosion monitoring data will depend on the number, location and variation in monitoring methods employed. For effective corrosion monitoring and control it is vital that all of the relevant data can be accessed easily, cross referenced and analysed. Typically in any production operation the data will be held on different databases and in a variety of formats (database or paper files). Therefore, it is essential to compile the relevant communication networks to facilitate this process. For example, BP Alaska have usefully integrated their corrosion and production databases. BPX Alaska are now using a Corrosion Analysis Tracking System (CATS). This computer system can store multi-giga-bytes of data from literally thousands of corrosion measurements and inspections in the field. The database is helping inspectors, corrosion engineers and

34

GENERAL GUIDELINES

others to develop a systematic, comprehensive approach to locating corrosion, analysing the best treatment strategies, and monitoring corrosion chemical treatments to verify their effectiveness. There are several proprietary database packages which can be used to do this. Typically they: ❍ ❍ ❍

generate monitoring and inspection reports generate inspection schemes, workscopes and plans demonstrate integrity status for certification purposes.

The main limitation with this type of package are the time taken to input data into the system and the lack of flexibility. However, such systems have the capability to become the main corrosion database for all the monitoring data. For example, CORTRAN (CORrosion TRend ANalysis) is currently used by two inspection contractors involved in the integrity management of the BP offshore assets in the UK sector of the North Sea [10]. Another albeit less efficient approach is to access all the databases and extract the relevant information needed. This is time consuming if undertaken manually and impacts on the effectiveness of any monitoring programme. Wytch Farm has developed a user friendly front end to their distributed control system [11]. This forms the management information system which archives and displays all site data for unlimited periods of time. It also has its own programming language which allows the user to develop high level applications e.g. energy monitoring or corrosion monitoring. The system also allows manual input of data such as the addition of laboratory reports or production engineering reports.

Another proprietary data handling and analysis package available is Mentor [12]. This system was developed for condition monitoring and has now been expanded to include corrosion monitoring information. The system can interrogate data from the distributed control system as well as data input manually. The software is such that the data can easily be compared from different databases and alarm levels set to alert an operator to potential changes in corrosivity. A “Mentor” system was installed on the Magnus asset in 1995.

35

GENERAL GUIDELINES

For effective analysis the following information is required: 1.

Process data: Usually available from a central database. This information involves both on-line and off-line data. These data should be supplemented by: ❍ ❍

Laboratory analysis Production engineering reports detailing - well shut-ins - acid stimulations - wireline activities - sand production - well workovers

2.

Corrosion monitoring data: Should include all the on-line data as well as the data collected manually (coupons etc.). These data should be stored in a format which enables direct comparison with the process data.

3.

Inspection data: Should include the routine inspection reports and data from specialised surveys. Again the data should be in a format which is directly comparable with the process and monitoring data.

The presentation of data is very important. The type of report is dependent on the activity and the scope of the work. Recently some operators have started to use CAD drawings as an aid to presenting corrosion and integrity data. This approach is very effective in identifying areas of concern and predicting potential locations of corrosion.

Side-stream Monitoring Side-stream monitoring is considered as a supplement to on-line corrosion monitoring. In this approach some of the process fluids are diverted from the facility into a temporary section of pipework containing the corrosion monitoring probes. The fluids then re-enter the main process stream or are collected for disposal later. This

36

GENERAL GUIDELINES

approach allows the flow rates to be modified and chemical treatments to be investigated without any major changes in production. Side-streams have been used extensively to study inhibitor performance. Although the use of side-streams appears to be useful there are several potential problems associated with their use. These are: ❍

The sampled fluids may not be representative of the process fluids.

❍

The side-stream may not simulate the correct flow regime for a given flow rate

❍

Side-streams tend to form well mixed fluids in low water cut situations therefore forming emulsions and preventing water separation.

❍

Temperature and pressure in the side-stream may not be representative of process stream.

In summary, side-streams should be used with caution and should never be used as a primary corrosion monitoring tool. Experience has shown them to be most effective on single phase systems (e.g. water injection flow lines). Any results obtained should be compared to field experience before reliance is placed on the results. BP Alaska used a side stream device to assess the performance of biocide in the sea water injection system. Biocide was terminated, based in part on the side stream data. Corrosion rates subsequently increase by approximately two orders of magnitude. Biocide was then restarted but even after 2 years it had not reduced the corrosion rates back to their previous levels

Corrosion Monitoring: A System by System Approach Background

This section serves to give examples of how the various monitoring techniques and approaches can be applied to given systems within an oil and gas production facility. These are only examples and in practice the monitoring system required may be quite different depending on the site specific conditions.

37

GENERAL GUIDELINES

Sea Water Injection Systems

Corrosion monitoring in sea water injection systems is quite complex and guidelines have been issued on corrosion control methods [13, 14] . Figure 10 summarises the basic corrosion monitoring required for a sea water system and the relevant references within this manual.

Figure 10: Corrosion Monitoring Requirements for a Sea Water Injection System

System

Monitoring

Corrosion Monitoring - Weight loss coupons - Electrical resistance methods - intrusive probe type - Flexible UT mata/auto UT - Electrochemical methodsb - intrusive probe type

Sea Water Injection systems

Process stream Monitoring - Flow rate - Temperature - Pressure - Iron counts - Dissolved oxygen (